Pharmacodynamic analysis and clinical trial of amoxicillin sprinkle administered once daily for 7 days compared to penicillin V potassium administered four times daily for 10 days in the treatment of tonsillopharyngitis due to Streptococcus pyogenes in children

Abstract

An a priori pharmacokinetic/pharmacodynamic (PK/PD) target of 40% daily time above the MIC (T >MIC; based on the MIC(90) of 0.06 microg/ml for Streptococcus pyogenes reported in the literature) was shown to be achievable in a phase 1 study of 23 children with a once-daily (QD) modified-release, multiparticulate formulation of amoxicillin (amoxicillin sprinkle). The daily T >MIC achieved with the QD amoxicillin sprinkle formulation was comparable to that achieved with a four-times-daily (QID) penicillin VK suspension. An investigator-blinded, randomized, parallel-group, multicenter study involving 579 children 6 months to 12 years old with acute streptococcal tonsillopharyngitis was then undertaken. Children were randomly assigned 1:1 to receive either the amoxicillin sprinkle (475 mg for ages 6 months to 4 years, 775 mg for ages 5 to 12 years) QD for 7 days or 10 mg/kg of body weight of penicillin VK QID for 10 days (up to the maximum dose of 250 mg QID). Unexpectedly, the rates of bacteriological eradication at the test of cure were 65.3% (132/202) for the amoxicillin sprinkle and 68.0% (132/194) for penicillin VK (95% confidence interval, -12.0% to 6.6%). Thus, neither antibiotic regimen met the minimum criterion of > or =85% eradication ordinarily required by the U.S. FDA for first-line treatment of tonsillopharyngitis due to S. pyogenes. The results of subgroup analyses across demographic characteristics and current infection characteristics and by age/weight categories were consistent with the primary-efficacy result. The clinical cure rates for amoxicillin sprinkle and penicillin VK were 86.1% (216/251) and 91.9% (204/222), respectively (95% confidence interval, -11.6% to -0.4%). The results of a post hoc PD analysis suggested that a requirement for 60% daily T >MIC(90) more accurately predicted the observed high failure rates for bacteriologic eradication with the amoxicillin sprinkle and penicillin VK suspension studied. Based on the association between longer treatment courses and maximal bacterial eradication rates reported in the literature, an alternative composite PK/PD target taking into consideration the duration of therapy, or total T >MIC, was considered and provides an alternative explanation for the observed failure rate of amoxicillin sprinkle.

FIG. 1.

Plot of percentages of T >MIC for individual subjects versus sequence number. Daily T >MIC for amoxicillin sprinkle QD was determined for individual subjects from a phase 1 study in pediatric patients with upper respiratory tract infections and was based on the MIC₉₅ (0.015 μg/ml) determined in the phase 3 clinical trial. The numbers of children receiving amoxicillin sprinkle were as follows: 6 months to 4 years old, 475 mg, fed, n = 11; 5 to 12 years old, 775 mg, fed, n = 12; 5 to 12 years old, 775 mg, fasted, n = 12.

FIG. 2.

Daily T >MIC was based on the MIC₉₅ of 0.015 μg/ml from the clinical trial and unbound drug concentrations in plasma. Daily T >MIC for amoxicillin sprinkle QD was determined from the results of a phase 1 study in pediatric patients with upper respiratory tract infections. Daily T >MIC for penicillin (Pen VK) QID was based on data in the literature (17). The numbers of children receiving amoxicillin sprinkle were as follows: 6 months to 4 years old, 475 mg, fed, n = 11; 5 to 12 years old, 775 mg, fed, n = 12; 5 to 12 years old, 775 mg, fasted, n = 12. Values displayed are the means ± standard deviations of the results.

FIG. 3.

PD endpoint assessment of projected failures based on different percentage of T >MIC PD endpoints (using phase 1 data) using unbound drug at a MIC of 0.015 μg/ml. The daily T >MIC data for amoxicillin sprinkle from a separate phase 1 study were fitted to a log normal distribution, and the log normal distribution parameters and 95% confidence intervals were determined (using JMP version 5.0.1a). The cumulative distribution of the fitted T >MIC data was then utilized to determine the projected number of therapeutic failures assuming a 40% T >MIC. The scale parameter estimate was 4.17, with a 95% confidence interval of 4.08 to −4.25. The shape parameter estimate was 0.25, with a 95% confidence interval of 0.21 to −0.31.

FIG. 4.

Plot of eradication rate versus T >MIC (assuming a MIC₉₅ of 0.015 μg/ml) achieved over the duration of therapy, i.e., total T >MIC, for various penicillin (Pen VK) dose regimens (▴) (4, 11, 12, 13, 14, 16, 18, 20, 21, 22, 23, 24, 25, 26, 30, 31, 32, 34, 37, 38, 39, 40, 41, 42, 43, 44, 45, 49, 50) compared to those achieved by amoxicillin sprinkle (▪, with 95% confidence interval shown as a bar) (based on average of both doses) and various amoxicillin dose regimens (♦) (1, 8, 11, 16, 33, 38) for the treatment of tonsillopharyngitis due to S. pyogenes. Model fit is based on penicillin VK data. The dashed line is at 85% eradication per U.S. FDA criterion standard.

FIG. 5.

Total T >MIC (days) was based on the MIC₉₅ of 0.015 μg/ml from the clinical trial and unbound-drug concentrations in plasma. Total T >MIC for amoxicillin sprinkle QD for 7 days was determined from the results of a phase 1 study in pediatric patients with upper respiratory tract infections. Total T >MIC for penicillin (Pen VK) QID for 10 days was based on data in the literature (17). The numbers of children receiving amoxicillin sprinkle were as follows: 6 months to 4 years old, 475 mg, fed, n = 11; 5 to 12 years old, 775 mg, fed, n = 12; 5 to 12 years old, 775 mg, fasted, n = 12. Values displayed are the means ± standard deviations of the results.